MNRAS 486, 3259–3265 (2019) doi:10.1093/mnras/stz1095

Laboratory formation and photo-chemistry of ionic HBC/anthraceneclusters in the gas phase

Junfeng Zhen,1,2‹ Weiwei Zhang,3 Yuanyuan Yang1,2 and Qingfeng Zhu1,2

1CAS Key Laboratory for Research in Galaxies and Cosmology, Department of Astronomy, University of Science and Technology of China,Hefei 230026, China2School of Astronomy and Space Science, University of Science and Technology of China, Hefei 230026, China3Department of Mechanical and Nuclear Engineering, Pennsylvania State University, University Park, PA 16802, USA

Accepted 2019 April 15. Received 2019 April 12; in original form 2019 February 4

ABSTRACTIn this paper, we present a study of the formation and photo-chemistry of hexa-peri-hexabenzocoronene (HBC, C42H18)/anthracene (C14H10) cluster cations in the gas phaseusing experimental and theoretical calculations, and we offer an approach of the evolutionof carbonaceous dust grains in the interstellar medium. The experiments, which combined aquadrupole ion trap and time-of-flight mass spectrometry, show that the large cluster cations canform by gas-phase condensation through molecular-ion reactions; that is, from C14H10DHBC+

(e.g. C14H10C42H17+, m/z = 699), to (C14H10)2DHBC+ [e.g. (C14H10)2C42H16

+, m/z = 876]and then to (C14H10)3DHBC+ [e.g. (C14H10)3C42H15

+, m/z = 1053]. Furthermore, with laserirradiation, the HBC/anthracene cluster cations involve a dehydrogenation process or the lossof its mono-anthracene groups. Based on the results of experiments and quantum chemistrycalculations, we demonstrate that large molecular clusters (∼100 atoms, covalently bonded)might be formed via a bottom-up process in the gas phase, and we show that there is a benefitto understanding their photo-chemical behaviour. In addition, the studies of HBC/anthraceneclusters (ranging from 83 to 129 atoms or ∼2 nm in size) also provide a way of interpretingnanometre-sized grains in space.

Key words: astrochemistry – molecular processes – methods: laboratory: molecular – ISM:molecules – ultraviolet: ISM.

1 IN T RO D U C T I O N

The mid-infrared spectra, observed in the interstellar medium(ISM) with broad features at 3.3, 6.2, 7.7, 8.6 and 11.2 μm, isoften recognized as the ultraviolet (UV) pumped fluorescence ofpolycyclic aromatic hydrocarbon (PAH) molecules (Sellgren 1984;Puget & Leger 1989; Allamandola, Tielens & Barker 1989). Manylines of evidence have shown that PAH clusters are very small dustgrains in the ISM (van Diedenhoven et al. 2004; Rapacioli, Joblin &Boissel 2005; Berne et al. 2007). Large PAH clusters or relatedcomplex species are believed to be formed by the self-assemblybetween free gas-phase PAHs and amorphous carbon particles.The evolution of the infrared (IR) spectrum suggests that the UVirradiation generally induces the destruction of fragile aliphaticbonds and the formation of a more aromatic system, which havebeen commonly thought of as an important role of the interstellarPAH model (Henning & Salama 1998; Rapacioli et al. 2006; Rheeet al. 2007; Tielens 2008).

� E-mail: jfzhen@ustc.edu.cn

Various studies on the photo-fragmentation of PAH moleculeshave been performed in the past decade (e.g. Ekern et al. 1998;Zhen et al. 2014; Castellanos et al. 2018; West et al. 2018). Par-ticularly, the photo-fragmentation behaviour of small PAH clustercations (∼50 atoms, e.g. pyrene cluster or fluorene cluster) hasbeen reported recently (Zhen, Chen & Tielens 2018; Zhang et al.2019). Compared with small PAH clusters, the formation processand photo-fragmentation behaviour of large PAH cluster cations(�100 atoms) are also very important in astrophysical applications,because they are initial forms of carbonaceous dust grains in theISM (Jager et al. 2009, 2011). However, there is still a lack ofexperimental evidence for these processes so far.

Carbonaceous grains are the most abundant cosmic dust com-ponents in space, and a major fraction of nanometre-sized car-bonaceous grains are formed by means of gas-phase condensation(Henning & Salama 1998). Such carbon particles are usually ablend of different carbon phases, including graphitic, diamond-like, fullerene-like and chain-like components on a sub-nanometreor nanometre scale (Kroto et al. 1985; Becker et al. 1994; vanDiedenhoven et al. 2004; Jager et al. 2011). In the dust model, largePAH clusters might play an important role in the formation pathways

C© 2019 The Author(s)Published by Oxford University Press on behalf of the Royal Astronomical Society

Dow

nloaded from https://academ

ic.oup.com/m

nras/article-abstract/486/3/3259/5480882 by University of Science and Technology of C

hina user on 05 May 2019

3260 J. Zhen et al.

of carbonaceous grains in space from molecular components orclusters (Henning & Salama 1998; Tielens 2008). Chen & Dobbins(2000) have proposed that large PAHs and PAH clusters with a layerstructure might be nucleation seeds.

In the PAH-dominated soot formation processes, at high tempera-tures, the driving force of the cluster growth is the generation of PAHradicals, which are formed by the release of hydrogen atoms fromPAHs (Violi & Izvekov 2007; Wang 2011; Elvati & Violi 2013).Then, the dominant pathway of the soot nuclei process is the reactionbetween PAH radicals and PAHs (Le Page, Snow & Bierbaum2001; Richter & Howard 2000; Richter et al. 2004). In this paper,we present a study of the formation process of large covalentlybonded PAH cluster cations (∼100 atoms), such as the hexa-peri-hexabenzocoronene (HBC, C42H18)/anthracene (C14H10) cluster,and their photo-fragmentation behaviour. Both experimental andtheoretical studies are carried out to improve our understanding ofthe inherent stabilities of large covalently bonded PAH clusters. Weexplore HBC/anthracene clusters using experiments with a newlyconstructed set-up – a quadrupole mass filter (QMF)/quadrupoleion trap source (QIT), analysed by time-of-flight (TOF) massspectrometry and supported by density functional theory (DFT)calculations. We select anthracene (C14H10, m/z = 178) for thisstudy, given its potential astrophysical interest (Tielens 2005) andthe presence of armchair and zigzag edges that might facilitate theformation of new molecule clusters. HBC (C42H18, m/z = 522) isan all-benzenoid PAH with a size in the astrophysically relevantrange (Tielens 2008; Croiset et al. 2016) that possibly serves asa prototypical example for large(r) PAHs. In addition, the photo-fragmentation behaviour and the IR spectra of HBC+ and HBC2+

have been obtained by Zhen et al. (2014, 2017).

2 EX P E R I M E N TA L ME T H O D S

We construct a new set-up to perform our experiment, a QMF–QIT–TOF mass spectrometer. In detail, the experimental apparatusconsists of three differentially pumped chambers. The first chamberis an ion source chamber (ionso chamber), which includes an oven(Heat Wave Labs) and an electron gun−ion gate system (EGUN,Jordan C-950). The second chamber is an ion trap chamber (iontrchamber) that contains a QIT (Jordan C-1251). The ionso and iontrchambers are connected by a QMF (Ardara), which consists offour 9-mm round rods, preceded by 1-inch pre-filter electrodes,and is fed by a radio-frequency (rf) power supply capable of anm/z in the range of up to 2000. The third chamber is detectionequipment, which is a reflection TOF mass spectrometer (JordanD-850). A 2-mm skimmer is placed between the iontr and TOFchambers. The base pressures in the ionso, iontr and TOF chambersare ∼3.0 × 10−8, ∼3.0 × 10−8 and ∼4.0 × 10−9 mbar, respectively.

HBC is evaporated by heating the powder (Kentax, with a puritybetter than 99.5 per cent) to the temperature of ∼573 K in theoven. Subsequently, the evaporated HBC molecules are ionizedby an EGUN and transported into the iontr chamber via the iongate and QMF. In this chamber, another oven (neutral moleculessource, ∼300 K) is located under the trap to make the vapourphase molecules (anthracene power, J&K, with a purity betterthan 99 per cent) effuse toward the centre of the ion trap, wherethe HBC/anthracene cluster cations are formed by the collisionreaction between cations and neutral anthracene molecules. Aftera short time delay (typically ∼0.2 s) with the ion gate closed,the stored waveform inverse Fourier transform excitation (SWIFT)pulse is applied to isolate species within a given mass/charge (m/z)range (Doroshenko & Cotter 1996). Helium gas is introduced

continuously into the trap to thermalize the ion cloud throughcollisions via a leak valve (Agilent). The working pressures in theionso, iontr and TOF chambers are ∼5.0 × 10−8, ∼6.0 × 10−7

and ∼7.0 × 10−9 mbar, respectively. Moreover, cluster cations arepresumably formed by collision stabilization of helium atoms underour experimental operating conditions. The third harmonic laserof Nd:YAG (INDI, Spectra-Physics), 355 nm, ∼6 ns, operated at10 Hz, 1.5 mJ (after the beam out the vacuum chamber, ∼3 mmin diameter, ∼2.5 × 105 W in peak power), is used to irradiate thetrapped species.

The typical frequency of 0.2 Hz is used for our set-up (i.e. onefull measuring cycle lasted 5 s), which benefits the accumulationof newly formed ion clusters. In our experiment, a beam shutter(Uniblitz, XRS−4) is used to act as a physical shield inside in theiontr chamber and determines the interaction time of the light andtrapped ion clusters. The shutter can only be externally triggeredto guarantee that the ion cloud is irradiated for a specified amountof time during each measurement cycle. We use a high-precisiondelay generator (SRS DG535) to control the full timing sequenceto make sure that each cycle begins with an empty ion trap. Theion gate opens (0–4.0 s) at the leading edge of the master trigger,allowed the ion trap to fill for a certain amount of ions. In thisperiod, the trapped ion collision reaction occurs between cationsand anthracene molecules to form new cluster cations. Once theclusters are formed, the SWIFT technique is applied to distinguisha specific range of m/z species (4.0–4.2 s). Afterwards (i.e. ∼0.2 s),the beam shutter opens, allowing the ion cloud to be irradiated (4.4–4.9 s). At the end of irradiation, a negative square pulse is appliedto the end cap of the ion trap to accelerate the ions out of the trapand transfer them into the field-free TOF region, where the massfragments can be measured. A LABVIEW program automates thefull data acquisition process. The resulting mass spectra are shownin Figs 1–4.

3 EXPERI MENTA L R ESULTS ANDDI SCUSSI ON

Fig. 1 shows a typical mass spectrum of HBC/anthracene clustercations, without either SWIFT isolation or laser irradiation. Aseries of peaks of HBC/anthracene cluster cations are observed,corresponding to different m/z species. Apart from the mainHBC mass peak (C42H18

+, m/z = 522), some mainly resid-ual HBC fragments are also measured, such as C42H14/15/16/17

+,m/z = 518, 519, 520, 521, namely as dehydrogenated HBC(DHBC), generated from the electron impact ionization (Zhenet al. 2014), and we only consider these fragments in the re-action pathway. In detail, the HBC/anthracene cluster cationsare labelled in a zoom-in mass spectrum in Fig. 1, showingthe new formed dehydrogenated HBC/anthracene cluster cations;that is, partially H-stripped ion species with the most abun-dant species, C14H10DHBC+ (e.g. C14H10C42H17

+, m/z = 699),(C14H10)2DHBC+ [e.g. (C14H10)2C42H16

+, m/z = 876] and(C14H10)3DHBC+ [e.g. (C14H10)3C42H15

+, m/z = 1053]. In ad-dition, we also observed one ‘extra’ peak (m/z = 791), although noassignments can be provided. We expect this peak might be formedas a side-product due to contaminations in the ion trap chamber.

The intensity ratio of cluster species is HBC+ :C14H10DHBC+ : (C14H10)2DHBC+ : (C14H10)3DHBC+ = 100 : 4 :0.6 : 0.1. This exponential decay of the intensity ratio demonstratesthat the collision formation reaction between DHBC and neutralanthracene molecules occurs step by step (Zhen et al. 2018;Zhang et al. 2019). According to our previous works (Zhen

MNRAS 486, 3259–3265 (2019)

Dow

nloaded from https://academ

ic.oup.com/m

nras/article-abstract/486/3/3259/5480882 by University of Science and Technology of C

hina user on 05 May 2019

Ionic HBC/anthracene clusters in the gas phase 3261

Figure 1. Mass spectrum of HBC/anthracene cluster cations trapped in the QIT before SWIFT isolation and without laser irradiation. The inset is a zoom-inmass spectrum, revealing the presence of newly formed clusters: C14H10DHBC+, (C14H10)2DHBC+ and (C14H10)3DHBC+.

Figure 2. Peak area of C14H10DHBC+ (without laser irradiation), with andwithout 13C correction.

et al. 2014, 2018), we compare the peak areas of (C14H10)HBCcluster cations with and without isotope (13C) correction, asshown in Fig. 2. It is clear that the m/z = 700 peak in Fig. 1is mainly the covalently bonded (C14H10)C42H17

+ cluster (i.e.13C12C55H27

+, m/z = 700), rather than the non-covalently bonded(C14H10)C42H18

+ (12C56H28+, m/z = 700).

Based on the observations, we propose the following formationreactions from C14H10DHBC to (C14H10)2DHBC, and then to(C14H10)3DHBC cluster cations. For the group of C14H10DHBC+

with different HBC dehydrogenated levels, we propose

C14H10 + C42H17+ → C14H10C42H17

+, (1)

C14H10 + C42H16+ → C14H10C42H16

+, (2)

C14H10 + C42H15+ → C14H10C42H15

+, (3)

C14H10 + C42H14+ → C14H10C42H14

+, (4)

Figure 3. Mass spectrum of HBC/anthracene cluster cations: with SWIFTisolation, but without (red) and with (blue) 1.5 mJ of laser irradiation(355 nm), and the differential spectrum (black).

for the group of (C14H10)2DHBC+ accordingly we propose

C14H10 + C14H10C42H16+ → (C14H10)2C42H16

+, (5)

C14H10 + C14H10C42H15+ → (C14H10)2C42H15

+, (6)

C14H10 + C14H10C42H14+ → (C14H10)2C42H14

+, (7)

and, finally, for the group of (C14H10)3DHBC+

C14H10 + (C14H10)2C42H15+ → (C14H10)3C42H15

+, (8)

C14H10 + (C14H10)2C42H14+ → (C14H10)3C42H14

+. (9)

Fig. 3 shows the mass spectrum of trapped HBC/anthracenecluster cations upon 355 nm irradiation with 1.5 mJ of laser energy(the entire irradiation time is 0.5 s, typically with ∼5 pulses),

MNRAS 486, 3259–3265 (2019)

Dow

nloaded from https://academ

ic.oup.com/m

nras/article-abstract/486/3/3259/5480882 by University of Science and Technology of C

hina user on 05 May 2019

3262 J. Zhen et al.

with SWIFT isolation. As we can see in Fig. 3 (upper black line,differential spectrum), the intensity of lower-mass peaks is increasedupon laser irradiation while the higher-mass peaks are decreasedfor each group of HBC/anthracene cluster cations. To understandthe photo-fragmentation behaviour of the HBC/anthracene clustercations in further detail, we illustrate the mass spectrum in Fig. 4(zoom-in and zoom-out of Fig. 3 in detail).

From Fig. 4(a) (upper black line, differential spectrum), forthe group of HBC cations, the intensity of newly formed species(lower-mass peaks, ‘profit’ ≈ 0.7) is more than the dissociationspecies (higher-mass peaks, ‘loss’ ≈ 0.1). However, for the groupof C14H10DHBC cluster cations (see Fig. 4b, upper black line,differential spectrum), the intensity of newly formed species (lower-mass peaks, ‘profit’ ≈ 0.8) is less than the dissociation species(higher-mass peaks, ‘loss’ ≈ 1.7) with laser treatment. This phe-nomenon indicates that, with laser irradiation, the C14H10DHBCcluster cations not only have a dehydrogenation channel but alsohave a mono-anthracene dissociation channel. For the intensity ofHBC+ and (C14H10)DHBC+, the ‘loss’ and ‘profit’ peaks in thedifferential spectra are nearly balanced, when we do not considerother peaks. The ‘extra’ number of HBC species are generatedfrom the dissociation of C14H10DHBC by the loss of its mono-anthracene group. The C14H10DHBC cations can either convert tomore aromatic species by the dehydrogenation process or shrinkthem back to small size molecules. by breaking the C–C bondbetween anthracene and HBC. Based on previous work by Zhenet al. (2018) and Zhang et al. (2019), we consider that to thedehydrogenation channel of HBC/anthracene cluster cations, wherethe H-atom is lost from anthracene (one carbon in sp3 hybrid orbitalwith a C–H bond).

In Figs 4(c) and (d) (upper black line, differential spectrum),for (C14H10)2DHBC and (C14H10)3DHBC cations, we cannot ob-serve any newly formed species, that might only have the mono-anthracene dissociation channel.

On the basis of the above discussion, we propose the followingphoto-dissociation channels upon laser irradiation for the threegroups of C14H10DHBC, (C14H10)2DHBC and (C14H10)3DHBCcluster cations: for the group of C14H10DHBC+

C14H10DHBC+ + hν → C14H9DHBC+ + H, (10)

C14H10DHBC+ + hν → DHBC+ + C14H10; (11)

for the groups of (C14H10)2DHBC+ and (C14H10)3DHBC+

(C14H10)2DHBC+ + hν → C14H10DHBC+ + C14H10, (12)

(C14H10)3DHBC+ + hν → (C14H10)2DHBC+ + C14H10. (13)

4 R E S U LT S O F T H E O R E T I C A LC A L C U L AT I O N S A N D D I S C U S S I O N

From the discussion of the mass spectrum, we have shown thata series of cationic clusters are generated in our experiment. Inthis section, we employ quantum chemistry calculations to linkthe experimental observations with their structures and possiblereaction pathways. The theoretical calculations are carried out withthe hybrid density functional B3LYP (Becke 1992; Lee, Yang &Parr 1988) as implemented in the Gaussian 16 program (Frischet al. 2016). The basis set of 6-311+ + G(2d,p) is used for allcalculations. To account for the intermolecular interaction, the

dispersion-correction (D3; Grimme, Ehrlich & Goerigk 2011) isalso considered for each system.

From the theoretical calculations, and as in Zhen et al. (2018),C14H10DHBC+ is a cationic cluster of mono-anthracene andHBC, connected by one C–C single bond. Here, the HBCis a dehydrogenated cation with a different dehydrogenatedlevel (i.e. C42H14/15/16/17

+). Accordingly, the cationic clusters are:[C14H10−C42H14]+ (m/z = 696); [C14H10−C42H15]+ (m/z = 697);[C14H10−C42H16]+ (m/z = 698); [C14H10−C42H17]+ (m/z = 699).For the C–C single bond in the C14H10DHBC+ cluster, one carbonfrom HBC is in sp2 hybridization, and the other carbon fromanthracene is in sp3 hybridization with an additional C–H bond.The formation reaction pathways of the C14H10DHBC group clustercations have been shown in equations (1)–(4).

As one typical example, in the following, we discuss the forma-tion pathway of [C14H10−C42H17]+, which is a cationic clusterconsisting of a neutral anthracene and C42H17

+. In Fig. 5(a),we present the possible reaction pathways of equation (1) andthe optimized structures of [C14H10−C42H17]+, from the singleC42H17

+ (positions A and B) and C14H10 (positions a, b and c)to the cationic cluster of [C14H10C42H17]+ (positions Aa, Ab, Ac,Ba, Bb and Bc). As a result, there might be six different possibleisomer structures for the cationic cluster. Clearly, all the reactionsare exothermic in the range from –2.7 to –3.6 eV, and the formationpathway with the lowest energy is C42H17

+ (B) + C14H10 (b) →[C14H10−C42H17]+ (Bb).

For the (C14H10)2DHBC group, as shown in Fig. 5(b), similarlyto C14H10DHBC+, the structure consists of two mono-anthracenemolecules and one DHBC, connected by two C–C single bonds:(C14H10)2C42H14

+, m/z = 874; (C14H10)2C42H15+, m/z = 875;

(C14H10)2C42H16+, m/z = 876. The DHBC fragment cations are

C42H14/15/16+, which at least have two dehydrogenated sites avail-

able. Equations (5)–(7) show the formation reaction pathway for(C14H10)2DHBC cations. For the (C14H10)3DHBC group clustercations, as shown in Fig. 5(c), similarly to C14H10DHBC+ and(C14H10)2DHBC+, the structure consists of three mono-anthracenemolecules and one DHBC, connected by three C–C single bonds:(C14H10)3C42H15

+, m/z = 1052 and (C14H10)3C42H16+, m/z = 1053.

The HBC fragment cations are (C42H14/15+), which as least have

three dehydrogenated sites available. Equations (8) and (9) showthe formation reaction pathway for (C14H10)3DHBC cations.

Because of the existence of too many isomer structures andreaction pathways, we only give one possible structure with onepossible reaction pathway (from B–B to B–Bb to Bb–Bb, andfrom B–B–B to B–B–Bb to B–Bb–Bb to Bb–Bb–Bb), as shownin Figs 5(b) and (c), which are based on the first step of thelowest energy reaction pathway (−3.6 eV): C42H17

+ (B) + C14H10

(b) → [C14H10−C42H17]+ (Bb). In our previous studies, we haveshown that the dissociation energy of H-loss is generally 2.0 eVfor aliphatic carbon (sp3) within PAH clusters (Zhen et al. 2018;Zhang et al. 2019), which is in line with the calculated covalentbond energy of these PAH cluster cations.

For PAH molecules, it is commonly accepted that excitationto an excited electronic state of a PAH molecule is followed byinternal conversion (IC) to a highly excited vibrational state of theground electronic state. Intramolecular vibrational redistribution(IVR) then quickly equilibrates the excess energy among allavailable vibrational states (IC and IVR occur on a time-scaleof picoseconds). In covalently bonded PAH cluster cations (e.g.DHBC/anthracene cluster cations), we suppose that the time-scaleof IC–IVR is similar to that of PAH molecules. In our experimentalconditions, the laser irradiation occurs on a nanosecond time-scale

MNRAS 486, 3259–3265 (2019)

Dow

nloaded from https://academ

ic.oup.com/m

nras/article-abstract/486/3/3259/5480882 by University of Science and Technology of C

hina user on 05 May 2019

Ionic HBC/anthracene clusters in the gas phase 3263

Figure 4. Mass spectrum of HBC/anthracene cluster cations, with SWIFT isolation, but without (red) and with (blue) 1.5 mJ of laser irradiation, and thedifferential spectrum (black): (a) in the range of m/z = 500–532 for HBC+; (b) in the range of m/z = 676–708 for C14H10DHBC+; (c) in the range ofm/z = 852–884 for (C14H10)2DHBC+; (d) in the range of m/z = 1028–1060 for (C14H10)3DHBC+.

(∼6 ns). This means that all the absorbed photon energy will beconverted into vibrational energy in the ground state, so we expectthat the dissociation process should take place in the electronicground state (Tielens 2008). As shown in Figs 5(b) and (c), thecovalent bond energy decreases with the size increase. For theC14H10DHBC cation, dehydrogenation (H-loss) is dominant. Forthe first dehydrogenation step in HBC/anthracene cluster cations,the C–H located in anthracene (carbon is in sp3 hybrid orbital) isfound to be the most favourable, and the dehydrogenation processof HBC/anthracene cluster cations is similar to that of pyrene dimercations (Zhen et al. 2018). Following dissociation initiated by thelaser irradiation, it is suspected that these covalently bonded clustersevolve towards either more aromatic species or cluster dissociation.Theoretical studies regarding the possible transition states anddynamical processes will be carried out in a subsequent paper(Zhang & Zhen, in preparation). In addition, further experimentswill be needed to address the evolution of collision reactions oflarge neutral PAHs with PAH cations.

5 A STRO NOMICAL IMPLICATIONS

The possible formation, stability and chemical role of complexesand (PAH) clusters in space is a very interesting topic, especiallywith regards to large molecular clusters (>100 atoms). It hasbeen suggested that the presence of PAH clusters can help to

explain certain characteristics of the astronomical observations ofthe unidentified IR emission spectra (Tielens 2013). Model studiesgenerally assume that these clusters are bonded by weak van derWaals forces (Potapov 2017). However, as pointed out, these bondsare very weak (Rhee et al. 2007), and such clusters might thus notsurvive a long time in the ISM (Rapacioli et al. 2006). Besides,recent experiments have shown us that the covalently bondedclusters of small PAHs can survive for a long time (even underconditions in which H-atoms are stripped off; Zhen et al. 2018;Zhang et al. 2019), even though the formation process of suchclusters is not entirely clear currently and hence it is still difficult toassess their formation in the ISM. This work provides further insightinto the evolution process of large covalently bonded PAH clusters(condensation). The evolution mechanism of DHBC/anthraceneclusters obtained here can be generalized, which can give us areasonable explanation for the photo-evolution of large molecularclusters and astronomically relevant and large PAHs (e.g. fromC14H10HBC to 56 C-atoms PAHs) in space (Andrews et al. 2015;Croiset et al. 2016).

Under laser irradiation, the large covalently bonded clusters willeither convert to their stable forms (i.e. form more aromatic speciesvia dehydrogenation processes) or shrink back to smaller PAHclusters or monomer species. Excitation of PAHs or PAH clustersin the ISM is mainly generated by single photon absorption. Asdiscussed earlier, in our experimental conditions all the absorbed

MNRAS 486, 3259–3265 (2019)

Dow

nloaded from https://academ

ic.oup.com/m

nras/article-abstract/486/3/3259/5480882 by University of Science and Technology of C

hina user on 05 May 2019

3264 J. Zhen et al.

Figure 5. The upper panel (a) is the formation reaction pathway for C14H10DHBC+ (e.g. C14H10C42H17+) using equation (1). The middle panel (b)

shows examples of fthe ormation reaction pathway for (C14H10)2DHBC+ [e.g. (C14H10)2C42H16+] using equations (2) and (5). The lower panel (c) is for

(C14H10)3DHBC+ [e.g. (C14H10)3C42H15+] using equations (3), (6) and (8).

photon energy is converted into vibrational energy in the groundstate, so the dissociation process should take place in the electronicground state (Tielens 2008). Hence, using our experimental results,for the photoprocessing of PAHs and PAH clusters in space, as inthe H I region (5.0 < E � 13.6 eV), it is suggested that the energyof a single UV photon is capable of achieving the photodissociationprocess of large PAH clusters. In the ISM, the size evolution isaccompanied by an evolution in the aliphatic-to-aromatic ratio(Pilleri et al. 2015; Zhen et al. 2018; Zhang et al. 2019). Thepresent experimental and theoretical studies address these links.However, irrespective of whether dehydrogenation or shrink backare involved, the direction of the photo-evolution of molecularclusters is always towards more aromatic species (Violi & Izvekov2007; Wang 2011; Candian, Zhen & Tielens 2018).

In addition, the structure of a pure carbon or hydrocarbon systemcan be more complicated in carbonaceous dust grains, as differenthybridized carbon (sp, sp2 and sp3) and mixed hybridization statescan occur in different ratios in one grain (Heidenreich, Hess &Ban 1968; Kroto et al. 1985; Jager et al. 2011). We stress thatgiven the different initial isomeric structures, from C14H10DHBC+

to (C14H10)2DHBC+ to (C14H10)3DHBC+, in the range from 83to 129 atoms or ∼2 nm in size, HBC/anthracene cluster is a goodexample of carbonaceous grains in space.

6 C O N C L U S I O N S

We have presented the first experiment on the formation andphoto-chemistry process of large covalently bonded PAH clusters(e.g. a family group of HBC/anthracene clusters, ∼100 atoms)in the gas phase. The formation process of the HBC/anthracenecluster cation provides a possible explanation for the evolutionof large molecular clusters in space. By studying the quantumchemical calculations, we demonstrate that the structure of theseHBC/anthracene clusters is diverse. Subsequent photoprocess-ing can further convert those clusters to larger PAHs or shrinkthem back to small monomer PAHs. Finally, HBC/anthraceneclusters, as an example of carbonaceous dust grains, provide abasic understanding of the grain evolution processes in the gasphase.

AC K N OW L E D G E M E N T S

This work is supported by the Fundamental Research Fundsfor the Central Universities and by the National ScienceFoundation of China (NSFC; Grant Nos 11421303 and11590782).

MNRAS 486, 3259–3265 (2019)

Dow

nloaded from https://academ

ic.oup.com/m

nras/article-abstract/486/3/3259/5480882 by University of Science and Technology of C

hina user on 05 May 2019

Ionic HBC/anthracene clusters in the gas phase 3265

REFERENCES

Allamandola L. J., Tielens A. G. G. M., Barker J. R., 1989, ApJS, 71, 733Andrews H., Boersma C., Werner M. W., Livingston J., Allamandola L. J.,

Tielens A. G. G. M., 2015, ApJ, 807, 99Becker L., Bada J. L., Winans R. E., Bunch T. E., 1994, Nature, 372, 507Becke A. D., 1992, J. Chem. Phys., 96, 2155Berne O. et al., 2007, A&A, 469, 575Candian A., Zhen J., Tielens A. G. G. M., 2018, Phys. Today, 71, 38Castellanos P., Candian A., Zhen J., Linnartz H., Tielens A. G. G. M., 2018,

A&A, 616, A166Chen H. X., Dobbins R., 2000, Combustion Science and Technology, 159,

109Croiset B. A., Candian A., Berne O., Tielens A. G. G. M., 2016, A&A, 590,

A26Doroshenko M. V., Cotter R. J., 1996, Rapid Communications in Mass

Spectrometry, 10, 65Ekern S. P., Marshall A. G., Szczepanski J., Vala M., 1998, J. Phys. Chem.

A, 102, 3498Elvati P., Violi A., 2013, Proc. Combust. Inst., 34, 1837Frisch M. J. et al., 2016, Gaussian 16 revision e. 01. Gaussian Inc.,

Wallingford, CTGrimme S., Ehrlich S., Goerigk L., 2011, Journal of Computational

Chemistry, 32, 1456Heidenreich R., Hess W., Ban L., 1968, Journal of Applied Crystallography,

1, 1Henning T., Salama F., 1998, Science, 282, 2204Jager C., Huisken F., Mutschke H., Llamas Jansa I., Henning T., 2009, ApJ,

696, 706Jager C., Mutschke H., Henning T., Huisken F., 2011, in Joblin C., Tielens

A. G. G. M., eds, EAS Pub. Ser. Vol. 46, PAHs and the Universe. EDPSciences, Les Ulis, p. 293

Kroto H., Heath J., O’Brien S., Curl R., Smalley R., 1985, Nature, 318, 162Le Page V., Snow T. P., Bierbaum V. M., 2001, ApJS, 132, 233Lee C., Yang W., Parr R. G., 1988, Phys. Rev. B, 37, 785

Pilleri P., Joblin C., Boulanger F., Onaka T., 2015, A&A, 577, A16Potapov A., 2017, Molecular Astrophysics, 6, 16Puget J. L, Leger A., 1989, ARA&A, 27, 161Rapacioli M., Joblin C., Boissel P., 2005, A&A, 429, 193Rapacioli M., Calvo F., Joblin C., Parneix P., Toublanc D., Spiegelman F.,

2006, A&A, 460, 519Rhee Y. M., Lee T. J., Gudipati M. S., Allamandola L. J., Head-Gordon M.,

2007, Proceedings of the National Academy of Science, 104, 5274Richter H., Howard J. B., 2000, Progress in Energy and Combustion Science,

26, 565Richter H., Granata S., Green W. H., Howard J. B., 2004, Proc. Combust.

Inst., 30, 1397Sellgren K., 1984, ApJ, 277, 623Tielens A. G. G. M., 2005, The Physics and Chemistry of the Interstellar

Medium, 1st edn. Cambridge Univ. Press, CambridgeTielens A. G. G. M., 2008, ARA&A, 46, 289Tielens A. G. G .M., 2013, Rev. Mod. Phys., 85, 1021van Diedenhoven B., Peeters E., Van Kerckhoven C., Hony S., Hudgins D.

M., Allamandola L. J., Tielens A. G. G. M., 2004, ApJ, 611, 928Violi A., Izvekov S., 2007, Proc. Combust. Inst., 31, 529Wang H., 2011, Proc. Combust. Inst., 33, 41West B., Castillo S. R., Sit A., Mohamad S., Lowe B., Joblin C., Bodi A.,

Mayer P. M., 2018, Physical Chemistry Chemical Physics, 20, 7195Zhang W., Si Y. B., Zhen J., Chen T., Linnartz H., Tielens A. G. G. M.,

2019, ApJ, 872, 38Zhen J., Paardekooper D. M., Candian A., Linnartz H. Tielens A. G. G. M.,

2014, Chem. Phys. Lett., 592, 211Zhen J., Castellanos, Bouwman J. D. M., Linnartz H., Tielens A. G. G. M.,

2017, ApJ, 836, 28Zhen J., Chen T., Tielens A. G. G. M., 2018, ApJ, 863, 128

This paper has been typeset from a TEX/LATEX file prepared by the author.

MNRAS 486, 3259–3265 (2019)

Dow

nloaded from https://academ

ic.oup.com/m

nras/article-abstract/486/3/3259/5480882 by University of Science and Technology of C

hina user on 05 May 2019